@inproceedings{zhou-etal-2026-parashard,
title = "{P}ara{S}hard: Knowledge Sharding for Effective Parametric Retrieval-Augmented Generation",
author = "Zhou, Changjiang and
Zhang, Ruqing and
Guo, Jiafeng and
Zhang, Yingchen and
Lv, Fuyu and
Cheng, Xueqi",
editor = "Lee, Hung-Yi and
Dev, Sunipa and
Li, Jessy and
Elliott, Desmond",
booktitle = "Findings of the {A}ssociation for {C}omputational {L}inguistics: {EMNLP} 2026",
month = oct,
year = "2026",
address = "Budapest, Hungary",
publisher = "Association for Computational Linguistics",
url = "https://aclanthology.org/2026.findings-emnlp.432/",
pages = "9084--9094",
ISBN = "979-8-89176-462-0",
abstract = "Retrieval-Augmented Generation (RAG) enhances large language models with external knowledge, but long retrieved contexts incur high computational cost and often degrade performance due to context{--}knowledge mismatch. Parametric RAG (PRAG) addresses this by integrating external knowledge into model parameters via lightweight adapters. However, existing PRAG methods are fundamentally document-centric, treating each document as an independent unit of knowledge, which leads to poor scalability and weak supervision. We argue that knowledge is inherently distributed across semantically related documents, and document-level parameterization is a suboptimal granularity. To this end, we propose ParaShard, a knowledge-centric framework that organizes knowledge into semantically coherent knowledge shards and parameterizes each shard with a shared adapter. This design enables richer supervision across related documents while significantly reducing the number of required adapters. At inference time, ParaShard dynamically selects and composes multiple shards using a diversity-aware strategy to ensure comprehensive knowledge coverage. Experiments on four knowledge-intensive QA benchmarks show that ParaShard achieves superior average performance across diverse knowledge-intensive tasks compared with strong RAG and PRAG baselines, with particularly strong gains on multi-hop reasoning, and substantially reduces storage cost compared to PRAG, demonstrating the knowledge injection effectiveness and storage efficiency of knowledge-centric parameterization."
}<?xml version="1.0" encoding="UTF-8"?>
<modsCollection xmlns="http://www.loc.gov/mods/v3">
<mods ID="zhou-etal-2026-parashard">
<titleInfo>
<title>ParaShard: Knowledge Sharding for Effective Parametric Retrieval-Augmented Generation</title>
</titleInfo>
<name type="personal">
<namePart type="given">Changjiang</namePart>
<namePart type="family">Zhou</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Ruqing</namePart>
<namePart type="family">Zhang</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Jiafeng</namePart>
<namePart type="family">Guo</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Yingchen</namePart>
<namePart type="family">Zhang</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Fuyu</namePart>
<namePart type="family">Lv</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Xueqi</namePart>
<namePart type="family">Cheng</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<originInfo>
<dateIssued>2026-10</dateIssued>
</originInfo>
<typeOfResource>text</typeOfResource>
<relatedItem type="host">
<titleInfo>
<title>Findings of the Association for Computational Linguistics: EMNLP 2026</title>
</titleInfo>
<name type="personal">
<namePart type="given">Hung-Yi</namePart>
<namePart type="family">Lee</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Sunipa</namePart>
<namePart type="family">Dev</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Jessy</namePart>
<namePart type="family">Li</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Desmond</namePart>
<namePart type="family">Elliott</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<originInfo>
<publisher>Association for Computational Linguistics</publisher>
<place>
<placeTerm type="text">Budapest, Hungary</placeTerm>
</place>
</originInfo>
<genre authority="marcgt">conference publication</genre>
<identifier type="isbn">979-8-89176-462-0</identifier>
</relatedItem>
<abstract>Retrieval-Augmented Generation (RAG) enhances large language models with external knowledge, but long retrieved contexts incur high computational cost and often degrade performance due to context–knowledge mismatch. Parametric RAG (PRAG) addresses this by integrating external knowledge into model parameters via lightweight adapters. However, existing PRAG methods are fundamentally document-centric, treating each document as an independent unit of knowledge, which leads to poor scalability and weak supervision. We argue that knowledge is inherently distributed across semantically related documents, and document-level parameterization is a suboptimal granularity. To this end, we propose ParaShard, a knowledge-centric framework that organizes knowledge into semantically coherent knowledge shards and parameterizes each shard with a shared adapter. This design enables richer supervision across related documents while significantly reducing the number of required adapters. At inference time, ParaShard dynamically selects and composes multiple shards using a diversity-aware strategy to ensure comprehensive knowledge coverage. Experiments on four knowledge-intensive QA benchmarks show that ParaShard achieves superior average performance across diverse knowledge-intensive tasks compared with strong RAG and PRAG baselines, with particularly strong gains on multi-hop reasoning, and substantially reduces storage cost compared to PRAG, demonstrating the knowledge injection effectiveness and storage efficiency of knowledge-centric parameterization.</abstract>
<identifier type="citekey">zhou-etal-2026-parashard</identifier>
<location>
<url>https://aclanthology.org/2026.findings-emnlp.432/</url>
</location>
<part>
<date>2026-10</date>
<extent unit="page">
<start>9084</start>
<end>9094</end>
</extent>
</part>
</mods>
</modsCollection>
%0 Conference Proceedings
%T ParaShard: Knowledge Sharding for Effective Parametric Retrieval-Augmented Generation
%A Zhou, Changjiang
%A Zhang, Ruqing
%A Guo, Jiafeng
%A Zhang, Yingchen
%A Lv, Fuyu
%A Cheng, Xueqi
%Y Lee, Hung-Yi
%Y Dev, Sunipa
%Y Li, Jessy
%Y Elliott, Desmond
%S Findings of the Association for Computational Linguistics: EMNLP 2026
%D 2026
%8 October
%I Association for Computational Linguistics
%C Budapest, Hungary
%@ 979-8-89176-462-0
%F zhou-etal-2026-parashard
%X Retrieval-Augmented Generation (RAG) enhances large language models with external knowledge, but long retrieved contexts incur high computational cost and often degrade performance due to context–knowledge mismatch. Parametric RAG (PRAG) addresses this by integrating external knowledge into model parameters via lightweight adapters. However, existing PRAG methods are fundamentally document-centric, treating each document as an independent unit of knowledge, which leads to poor scalability and weak supervision. We argue that knowledge is inherently distributed across semantically related documents, and document-level parameterization is a suboptimal granularity. To this end, we propose ParaShard, a knowledge-centric framework that organizes knowledge into semantically coherent knowledge shards and parameterizes each shard with a shared adapter. This design enables richer supervision across related documents while significantly reducing the number of required adapters. At inference time, ParaShard dynamically selects and composes multiple shards using a diversity-aware strategy to ensure comprehensive knowledge coverage. Experiments on four knowledge-intensive QA benchmarks show that ParaShard achieves superior average performance across diverse knowledge-intensive tasks compared with strong RAG and PRAG baselines, with particularly strong gains on multi-hop reasoning, and substantially reduces storage cost compared to PRAG, demonstrating the knowledge injection effectiveness and storage efficiency of knowledge-centric parameterization.
%U https://aclanthology.org/2026.findings-emnlp.432/
%P 9084-9094
Markdown (Informal)
[ParaShard: Knowledge Sharding for Effective Parametric Retrieval-Augmented Generation](https://aclanthology.org/2026.findings-emnlp.432/) (Zhou et al., Findings 2026)
ACL